RFR: 8262896: [macos_aarch64] Crash in jni_fast_GetLongField

Tue Apr 13 12:24:57 UTC 2021

On Tue, 13 Apr 2021 12:11:55 GMT, Andrew Haley <aph at openjdk.org> wrote:

> It depends on performance. I've had a quick look at how `pthread_jit_write_protect_np` works, and it looks like some serious overhead. Can you measure, say, a few million JNI accessors to see which works best?

I just had a look, and according to https://dougallj.github.io/applecpu/firestorm.html `MSR (APRR)` only takes a single cycle, but there is still some overhead because of an `ISB`. Please try it and see, though.

-------------

PR: https://git.openjdk.java.net/jdk/pull/3422